This project explores how digital lifeforms can be evolved and shaped through natural language guidance. By integrating CLIP-based semantic evaluation with CMA-ES optimization, users can co-evolve artificial life based on prompts such as “a peaceful bird” or “a fierce jellyfish.” The interface supports real-time interaction and fosters an embodied sense of intention and emergence in artificial life design.
We also incorporated swarm behaviors and interactive feedback loops to allow participatory experimentation with co-creative agents. The work aims to bridge computational intelligence, generative aesthetics, and the future of embodied design.
← Back to Home